Comment by jackero
12 hours ago
I legitimately tried my best to like XLST on the web back in the day.
The idea behind XLST is nice — creating a stylesheet to transform raw data into presentation. The practice of using it was terrible. It was ugly, it was verbose, it was painful, it had gotchas, it made it easier for scrapers, it bound your data to your presentation more, and so on.
Most of the time I needed to generate XML to later apply a XLST style sheet, the resulting XML document was mostly a one off with no associated spec and not a serious transport document. It begged the question of why I was doing this extra work.
Making your data easy to scrape is part of the point (or just more generally work with). If you're building your web presence, you want people to easily be able to find the data on your site (unless your goal is platform lockin).
The entire point of XSLT is to separate your data from its presentation. That's why it made it easy to scrape. You could return your data in a more natural domain model and transform it via a stylesheet to its presentation.
And in doing so it is incredibly concise (mostly because XPath is so powerful).